Yasushi Kawakura

Improving the performance of global communication on a three‐dimensional torus network

A high-speed one-to-all broadcasting algorithm is proposed whose performance does not deteriorate much when the number of processors is increased in a massively parallel computer. For the network topology, 3D torus networks are considered. Two methods are discussed for a system which broadcasts by repeating one-to-one communications. One uses paths having a smaller maximum transfer number to reduce the number of transfers, and the other presets the hardware to reduce the overhead of individual one-to-one communications. These methods are evaluated using a double loop model which consists of an inner loop for local processing and an outer loop for global communications. When these methods are used, the scalability increases and for a 32K processor system a 4.2 times speedup in program execution can be achieved.

Improving the Performance of Global Communication on a 3D Torus Network

The authors developed an one-to-all broadcasting algorithm which is less dependent on the increase of processors, by reducing the overhead of each communications, and by shortening the message transmission sequence. The result shows that the program which employs this algorithm can achieve a 2.8 times speedup in the case of a 32K processor system.